README "Replication Data for: Remedial Education: Evidence from a Sequence of Experiments in Colombia." (By Alvarez Marinelli, H., S. Berlinski and M. Busso)


Files

1. "datawide.dta": This dataset includes the variables used for Table 1. It is in wide format, where each row is a student.

2. "datalong.dta": This dataset includes the variables used in most tables and figures of the paper (except for Table 1 and Table 6). It is in long format (at the student-time level, where time is beginning of Grade3, end of Grade 3, beginning of Grade 4, end of Grade 4). 

Because combinations of the raw (non-standardized) scores would allow to identify students'/schools' treatment status (by someone with access to the scores data) we cannot release the raw scores. For this reason, the data to replicate Table 6 is not available. Please contact the corresponding author for inquiries. As a substitute we publish aggregate statistics (means and s.d.) of those raw scores by treatment status for each time of measurement. 

3. "scorestabs.dta": This dataset presents aggregate statistics (mean and standard deviation) for non-standardized scores for each cohort-treatment status.

4. "scripts.zip": This zip file contains the main dofile that replicates the tables and figures of the paper using as input previously presented datasets. Please see notes at the beginning of each table/figure in dofile for more information. It also contains auxiliary ado files that are called by the main docile.
